Crawl of the Friendster social network

During my master thesis research, I did a web crawl of the Friendster social network.

Data format

It's a gzipped relations text file with on each line:

[user-id1],,[username1]->[user_id2],,[username2]

This means that user1 with user_id1 and username1 has user2 as friend (an edge between user1 and user2). The friendster graph has been crawled by storing directed relations. This means that both user1->user2 and user2->user1 may be in de relations file.

General information

Further information about the crawl:

• Site: www.friendster.com
• Relations(directed): 28 million
• Bootstrap user: 9646373 username: Joy
• Begin crawl date: Thu Nov 17 15:13:31 2005
• End crawl date: Wed Dec 7 15:55:05 2005

Using the data

If you like to use the crawl data, please contact me or the Tribler team.

• Link to the crawl data: Friendster crawl data

Reference

If you have used the data in any publication, please refer to my Research assigment (unpublished):

J. Roozenburg. A literature survey on bloom filters. Research
Assignment, November 2005.

• BIB:

  @UnPublished{ bloomfilter_jelle,
   author =       {J. Roozenburg},
   title =        {A literature survey on Bloom filters},
   school =       {Delft University of Technology},
   year =         2005,
   address =      {Delft, NL},
   month =        {November},
   note =         {Research Assignment}
  }